Changes in force and cytosolic Ca²⁺ concentration after length changes in isolated rat ventricular trabeculae

Abstract

1. Changes in cytosolic [Ca²⁺] ([Ca²⁺]_i) were measured in isolated rat trabeculae that had been micro-injected with fura-2 salt, in order to investigate the mechanism by which twitch force changes following an alteration of muscle length.

2. A step increase in length of the muscle produced a rapid potentiation of twitch force but not of the Ca²⁺ transient. The rapid rise of force was unaffected by inhibiting the sarcoplasmic reticulum (SR) with ryanodine and cyclopiazonic acid.

3. The force-[Ca²⁺]_i relationship of the myofibrils in situ, determined from twitches and tetanic contractions in SR-inhibited muscles, showed that the rapid rise of force was due primarily to an increase in myofibrillar Ca²⁺ sensitivity, with a contribution from an increase in the maximum force production of the myofibrils.

4. After stretch of the muscle there was a further, slow increase of twitch force which was due entirely to a slow increase of the Ca²⁺ transient, since there was no change in the myofibrillar force-[Ca²⁺]_i relationship. SR inhibition slowed down, but did not alter the magnitude of, the slow force response.

5. During the slow rise of force there was no slow increase of diastolic [Ca²⁺]_i, whether or not the SR was inhibited. The same was true in unstimulated muscles.

6. We conclude that the rapid increase in twitch force after muscle stretch is due to the length- dependent properties of the myofibrils. The slow force increase is not explained by length dependence of the myofibrils or the SR, or by a rise in diastolic [Ca²⁺]_i. Evidence from tetani suggests the slow force responses result from increased Ca²⁺ loading of the cell during the action potential.

It has been known since the work of Parmley & Chuck (1973) that a reduction in length of isolated cardiac muscle produces a rapid decrease of twitch force, followed by a further, slow fall of force over 10 min or so. If length is increased, there are corresponding rapid and slow increases in twitch force (reviewed by Allen & Kentish, 1985; Lakatta, 1992). The rapid changes in force are thought to form the basis of the length-tension (Frank-Starling) mechanism in the whole heart. The slow force responses to a length change are also likely to be important for the regulation of cardiac function, since they have been seen in blood-perfused isolated hearts (e.g. Tucci, Bregagnollo, Spadaro, Cicogna & Ribeiro, 1984; Burkhoff, de Tombe, Hunter & Kass, 1991) and in vivo in anaesthetized dogs (Lew, 1993). Indeed, in vivo they probably underlie the ‘Anrep effect’, by which an increase in ventricular volume due to a rise in aortic pressure is followed by a secondary, slow increase in myocardial performance, such that the ventricular volume falls towards its original size. However, the mechanism of the slow force responses has remained obscure. Allen & Kurihara (1982) showed that the slow increase in force after stretch of isolated papillary muscles was due, at least in part, to a slow rise in the magnitude of the intracellular Ca²⁺ transient. This slow enhancement of the Ca²⁺ transient could have been due to a potentiation of excitation-contraction coupling, or it could have risen indirectly by a rise in diastolic Ca²⁺ concentration, which would then lead to greater Ca²⁺ sequestration by the sarcoplasmic reticulum (SR). Subsequently, Nichols (1985) reported that diastolic length was the important factor controlling the slow force response to a length change in papillary muscles. Thus it seemed likely that stretch did increase the diastolic Ca²⁺ concentration, perhaps by activation of the stretch-activated channels reported in cardiac cells by a number of workers (e.g. Craelius, 1993). However, it has proved difficult to establish whether diastolic Ca²⁺ concentration does indeed increase after muscle stretch. Allen, Nichols & Smith (1988) used aequorin to confirm that it was the diastolic length of papillary muscles that controlled the size of the systolic Ca²⁺ transient, but the insensitivity of aequorin to resting levels of Ca²⁺ made it difficult to measure the diastolic Ca²⁺ concentration. More recent studies have used fura-2, which has the appropriate K_d to measure diastolic levels of Ca²⁺ (∼200 nM; Grynkiewicz, Poenie & Tsien, 1985). In a preliminary study, Steele & Smith (1993) reported that the diastolic Ca²⁺ concentration in guinea-pig trabeculae did increase during the slow force responses. In contrast, Hongo, White, Le Guennec & Orchard (1996) found diastolic Ca²⁺ concentration was unchanged during the slow force response in rat myocytes. In addition to this discrepancy, in both studies the preparations were loaded with the acetoxymethyl (AM) form of fura-2, which has the disadvantage that fura-2 AM enters, and gives a fluorescence signal from, intracellular organelles such as mitochondria. In addition, there can be a Ca²⁺-independent fluorescence signal from partly hydrolysed fura-2 AM. Both factors may obscure true changes in the Ca²⁺ concentration of the cytosol (Backx & ter Keurs, 1993).

Another source of uncertainty is that it is not known whether the magnitude of the slow force responses can be attributed entirely to changes in the systolic Ca²⁺ transient, or whether there is also a contribution from a change in the Ca²⁺ sensitivity of the myofibrils.

In the present work, we investigated the influence of muscle length on the cytosolic Ca²⁺ concentration ([Ca²⁺]_i) during systole and diastole in rat trabeculae. We iontophoresed fura-2 salt into the myocardial cells of the trabeculae, using the technique of Backx & ter Keurs (1993). This procedure not only allows an unequivocal measure of cytosolic [Ca²⁺]_i alone, but also ensures that the fura-2 signal comes entirely from the myocardial cells and not from the smooth muscle and endothelial cells in the preparation. In addition, we determined the force-[Ca²⁺]_i relationship for the myofibrils in situ in the trabeculae, and so could determine, for the first time, whether changes in this relationship contributed to the slow force responses. Finally, we examined the relative contributions of the length-dependent changes in maximum force production and Ca²⁺ sensitivity of the myofibrils to the rapid effects on twitch force of a change in muscle length.

A preliminary account of this work has been presented (Kentish & Wrzosek, 1995).

METHODS

Preparation of trabeculae

Male Wistar rats (∼250 g) were stunned by a blow to the head and were killed by cervical dislocation under Home Office guidelines (Schedule 1). Their hearts were removed and washed free of blood with Tyrode solution containing (mM): NaCl, 93; NaHCO₃, 20; Na₂HPO₄, 1; MgSO₄, 1; KCl, 5; CaCl₂, 1; glucose, 10; sodium acetate, 20; insulin, 5 U l⁻¹; oxygenated with 95% O₂-5% CO₂; pH 7.4 at room temperature (23°C). Trabeculae (1.5–3 mm long, 90–150 μm wide and 50–100 μm thick), with a small piece of mitral valve attached, were dissected from the right ventricle in Tyrode solution containing 25 mM 2,3-butanedione monoxime (BDM) to minimize muscle damage (Mulieri, Hasenfuss, Ittleman, Blanchard & Alpert, 1989). A trabecula was then mounted in a perfusion bath (5 mm × 4 mm × 5 cm) located on a stage of a Nikon Diaphot inverted microscope and was superfused with Tyrode solution. The valve end of the muscle was impaled on a fine hook connected to a force transducer (SensoNor, Horten, Norway) and the wall end was ensnared in a wire loop attached to a Narishige micromanipulator. A digitimer D4030 (Digitimer Ltd, Welwyn Garden City, Herts, UK) triggered an isolated stimulator (DS2, Digitimer), which delivered pulses (5 ms; 10–20% above threshold) at 0.33 Hz to the muscles via platinum electrodes running along each side of the muscle bath. During an equilibratory period of > 1 h the muscles were stretched progressively to the length where developed (‘twitch’) force was ∼95% of its maximum. This initial length is termed L_0. Muscles were shortened, usually to 90% of L₀ (L₉₀), by manual adjustment of the micromanipulator (taking 3–5 s to complete).

All experiments were done at room temperature (23–25°C). For unstimulated muscles and for twitches we used an extracellular [Ca²⁺] ([Ca²⁺]_o) of 1 mM, which is close to the value of 1.3 mM for the ionized [Ca²⁺]_o in rat blood (Forester & Mainwood, 1974; Chambers, Braimbridge & Hearse, 1991). For tetanic stimulation [Ca²⁺]_o was raised to 8 mM to ensure maximum Ca²⁺ activation of the myofibrils.

Cell loading with fura-2

Before injection of fura-2, the 340 and 380 nm autofluorescence signals at muscle lengths L₀ and L₉₀ were measured. Fura-2 pentapotassium salt was then iontophoresed into myocardial cells in the trabecula using the method of Backx & ter Keurs (1993) with small modifications. Briefly, the tip of a microelectrode (final resistance, 50–300 MΩ) was filled with 2 mM fura-2 pentapotassium salt in H₂O and was backfilled with 140 mM KCl. The perfusion solution was changed to Tyrode solution containing 10 mM BDM and stimulation was stopped. Then the microelectrode was advanced into a cell, which usually had a membrane potential of −55 to −80 mV, measured using an AxoClamp-2A amplifier (Axon Instruments, Foster City, CA, USA). When the voltage was stable, a negative current of 4–8 nA was passed for 8–15 min (depending on the size of the muscle) to inject fura-2 into the cell. The muscle was then superfused with normal Tyrode solution and fura-2 fluorescence was measured. Usually two or three injections were given at different sites along the muscle in order to ensure that fura-2 fluorescence was 3–5 times above autofluorescence and to increase the uniformity of dye distribution. After iontophoresis, the muscle was stimulated at 0.33 Hz for about 1 h, during which fura-2 spreads along the muscle via the gap junctions (Backx & ter Keurs, 1993). We measured dye distribution by recording the 340 and 380 nm fluorescence signals through a small (300–450 μm square) window that was moved along the muscle. Figure 1 shows an example of a muscle in which we gave three injections. The uniformity of dye distribution over the typical recording distance for the length-change experiments (∼1 mm) was usually > 80% after 1 h (Fig. 1), when the experiments were started. The dye uniformity was increased by the continued diffusion of dye during the experiment, although dye concentration fell due to loss of fura-2 from the cells (Fig. 1). Any slight non-uniformity of dye distribution should not affect the results because we measured the 340 nm/380 nm fluorescence ratio, which is independent of dye concentration.

Figure 1. Distribution of fura-2 along an isolated rat trabecula at various times after iontophoresis of fura-2 into three myocardial cells in the preparation.

A, absolute fluorescence readings. The distribution of fura-2 fluorescence plus muscle autofluorescence (excitation, 380 nm) was recorded in a ≈450 μm-wide window passed along the muscle at the following times after the end of the iontophoresis: •, 30 min; ○, 60 min; ▪, 120 min; and □, 210 min. Arrows indicate the approximate sites of iontophoresis. Distance was measured from the start of muscle tissue at the valvular end of the muscle. Similar results were recorded with 340 nm excitation (not shown). B, same data as in A, but expressed relative to the maximum fluorescence at that time. The horizontal arrow shows the typical recording distance (≈1 mm) used in the length-change experiments. The relative rise of fura-2 fluorescence at the right-hand side later in the experiment occurred as fura-2 diffused into the greater mass of tissue at this (ventricular wall) end of the muscle.

Over the period required for successful multiple injections and post-injection equilibration (2–3 h), twitch force declined by about 20%. Although some of this decrease may have been due to buffering of cytosolic [Ca²⁺]_i by the fura-2 (see below), most of it was probably caused by the tendency of the twitch in rat ventricular muscle to decline throughout an experiment (e.g. see Allen & Kurihara, 1982; J. C. Kentish & A. Wrzosek, unpublished observations).

Fluorescence measurements

The ratio of fura-2 fluorescence at 510 nm with excitation at 340 and 380 nm was measured with a Cairn spectrophotometer (Cairn Research, Faversham, Kent, UK). Light from a 75 W xenon arc lamp was filtered using 340 and 380 nm bandpass filters (bandwidth, 10 nm) in a six-filter wheel rotating at ∼80 Hz. This excitation light was passed through a liquid light guide to the microscope, was reflected by a 400 nm dichroic mirror and was focused onto the muscle via × 10 objective (Nikon UV-Fluor). The excitation shutter was opened only for brief recording periods (usually 48 s, to record 16 twitches) to minimize photobleaching of fura-2. About 1.4 mm (i.e. 50 −90%) of the trabecula was illuminated by the excitation light. The image of the muscle was also observed using red illuminating light (> 700 nm) and a video camera. The fluorescence light was collected by the objective and passed through a 510 nm filter (bandwidth, 50 nm) to the photomultiplier. The Cairn circuitry subtracted the autofluorescences from the 340 and 380 nm fura-2 fluorescence signals and calculated the running mean of the 340 nm/380 nm fluorescence ratio (equivalent to a mean at 80 Hz). Force and fluorescence signals were recorded on a chart recorder (with filtering at 15 Hz) and were averaged and stored (without filtering) on a computer using a Digidata 1200 A/D board and pCLAMP software (Axon Instruments).

Diastolic fluorescence ratio was measured as the mean value over a 100 ms period just before the stimulus. The systolic fluorescence ratio was determined directly from the peak of the recorded ratio or, if the Ca²⁺ transients were noisy, from the peak of a bi-exponential fit to the Ca²⁺ transient. The steady-state systolic fluorescence ratio at L₀ was found to increase slowly over the course of the experiments (Fig. 3B), and was associated with a progressive shortening of the decay of the Ca²⁺ transient (Fig. 4D and F). These changes have been explained by Backx & ter Keurs (1993) as being due to the slow loss of fura-2 from cells, leading to less buffering of [Ca²⁺]_i by fura-2. However, the statistical results shown in the figures were the same whether or not we corrected the fluorescence ratios for this decline (assuming a linear decline with time). We did not undertake the lengthy calibration procedure (Backx & ter Keurs, 1993) needed to convert from fluorescence to [Ca²⁺]_i because our primary interest was in the relative changes in [Ca²⁺]_i after changes of length.

Figure 3. The effects of the length changes on twitch force and on the diastolic and systolic fluorescence ratios (340 nm/380 nm).

A, twitch (active) force during the different periods (1–5) of Fig. 1A. B, systolic fluorescence ratio (○) and diastolic ratio (▵) for the same periods. Note the break in the ordinate. The abscissae indicate muscle length and the time in minutes at which data recording (lasting 48 s) was started after the length change. L₉₀ represents a length of 90%L₀. Points show means ±s.e.m., n = 15. Where no error bars are shown, s.e.m. is less than the size of the symbol. Symbols next to the data points indicate results of paired t tests between each value and the corresponding value in the preceding period. *P < 0.05; †P < 0.001.

Figure 4. The effects of the length changes on the time course of the twitch and fluorescence transients.

Left-hand panels, force transients; right-hand panels, 340 nm/380 nm fluorescence transients. A and B, time from the stimulus to peak force and fluorescence. C and D, time from the peak to 50% relaxation (RT₅₀) of force and fluorescence. E and F, time from the peak to 90% relaxation (RT₉₀) of force and fluorescence. The abscissae indicate muscle length and the time in minutes at which data recording was started after the length change. L₉₀ represents a length of 90%L₀. Data are expressed as means ±s.e.m., n = 15. Symbols next to the data points indicate results of paired t tests between each value and the corresponding value in the preceding period. *P < 0.05; †P < 0.01.

In these experiments the Cairn circuitry subtracted constant values for autofluorescences at 340 and 380 nm (measured at L₀) from the fura-2 fluorescence signals before calculation of the 340 nm/380 nm fluorescence ratio. It is therefore important to consider whether autofluorescence might have changed, and caused artifacts in the calculated 340 nm/380 nm ratio, under the various conditions of the experiment.

During the twitch

Muscles that showed substantial autofluorescence changes during the twitch, due to shortening of the central section, were discarded. In the muscles that were judged suitable for these experiments there was little or no autofluorescence change during the twitch.

Immediately after a length change

Shortening the muscles by 10% from L₀ to L₉₀ increased autofluorescences at 340 and 380 nm by 10.8 ± 2.7% and 9.3 ± 2.3%, respectively (means ±s.e.m., n = 8), as expected from the increase of muscle volume visible to the photomultiplier. We calculate that subtraction of the autofluorescence measured at L₀ from the fura-2 signals measured at L₉₀ typically would have lead to our underestimating the diastolic 340 nm/380 nm ratio at L₉₀ by ∼5% and overestimating the systolic ratio at L₉₀ by ∼4%. However, this would not alter our conclusions regarding the major focus of our paper - the slow changes in [Ca²⁺]_i - because these slow changes occurred with the muscle held at a constant length.

During the slow force responses to a length change

Autofluorescence arises mainly from mitochondrial NAD(P)H, and might be expected to vary during the slow changes of force. In fact we found that autofluorescence was not altered during the slow force responses. For example, over the 15 min period after re-lengthening of the muscle the autofluorescences at 340 and 380 nm changed by 1.5 ± 0.9% and 2.1 ± 0.7% (n = 4), respectively, which were not significantly different from zero (P > 0.05, Student's paired t test). Thus our calculations of 340 nm/380 nm ratios during the slow force responses are not complicated by changes in autofluorescence.

Upon cessation of stimulation

Stopping stimulation (0.33 Hz) increased the autofluorescence at 340 nm by 5.5 ± 1.2% (n = 5; P < 0.05), but did not affect the 380 nm autofluorescence (increase of 1.5 ± 0.9%; P > 0.05). We did not correct for the 340 nm change, which would have reduced the 340 nm/380 nm ratios in the quiescent muscles (Fig. 7) by ∼1%. This is a negligible error and does not alter any of our conclusions from the results in Fig. 7.

Figure 7. Fura-2 fluorescence ratio after length changes in unstimulated muscles.

A, in normal muscles with active SR (n = 4). B, in muscles treated with 1 μM ryanodine plus 30 μM cyclopiazonic acid to inhibit the SR (n = 3). The abscissae indicate muscle length and the time when data recording was started after the length change. L₉₀ represents a length of 90%L₀. Points show mean values ±s.e.m.

Data analysis

Force and fluorescence were usually averaged over 48 s periods, beginning immediately after, or 15 min after, the length change. The mean responses over 48 s after the length change are defined as ‘rapid’ responses of force or fluorescence due to the length change, whereas the subsequent changes to the 15 min (quasi-steady-state) condition are referred to as ‘slow’ responses. Results are expressed as means ±s.e.m. Statistical analysis was performed by Student's paired or unpaired t tests, as appropriate.

Chemicals

Fura-2 was from Molecular Probes, Inc. (Eugene, USA) and cyclopiazonic acid (CPA) and ryanodine were from Calbiochem (La Jolla, USA). CPA was added to the Tyrode solution from stock solution dissolved in DMSO; the DMSO (final concentration 0.1%) did not by itself affect twitch force. Other chemicals were analytical grade from Sigma (Poole, Dorset, UK).

RESULTS

Force and [Ca²⁺]_i in muscles with functional SR

Typical changes in force and fura-2 fluorescence ratio (reflecting the cytosolic [Ca²⁺]) produced by a 15 min decrease in muscle length are shown in Fig. 2. In this muscle the 10% decrease in muscle length from L₀ to L₉₀ immediately lowered twitch force by 70%, with a small decrease in resting force, and this was followed over the next 15 min by a further fall of twitch force. After re-lengthening the muscle to L₀, there was a rapid, then a slow rise of twitch force to its original level. Thus rat trabeculae showed both the rapid and slow force responses to a length change that have been seen in a number of other myocardial preparations. The fura-2 fluorescence ratio (Fig. 2B) also exhibited length- and time-dependent changes; these are most clearly seen from overlaid traces illustrating the rapid changes (Fig. 2C) and the slow changes (Fig. 2D) in force and fluorescence after re-lengthening of the muscle. These mean traces illustrate the high signal-to-noise ratio produced by recording the fluorescence from hundreds of cells in the trabeculae; this allowed us to detect very small changes in cytosolic [Ca²⁺] after changes in muscle length. The rapid increase in twitch force occurred with virtually no change in the magnitude of the Ca²⁺ transient (Fig. 2C), but the slow increase in force was associated with a rise of the Ca²⁺ transient (Fig. 2D). Mean data for the entire release/re-stretch protocol in fifteen muscles are presented in Fig. 3. Although there was some variability of absolute force and fluorescence ratio between muscles, the relative changes in each muscle were very consistent. The slow fall of force following muscle shortening to 90%L₀ (i.e. from periods 2 to 3), and the slow rise of force after re-lengthening of the muscle (from periods 4 to 5), were associated with highly significant (P < 0.001) changes in the systolic fluorescence ratio (Fig. 3B). In contrast, there were no significant rapid changes of Ca²⁺ transient magnitude after shortening or lengthening of the muscles. We found that the relative changes in force and fluorescence after the muscle was shortened were almost always equal and opposite to those seen when the muscle was re-lengthened. Therefore, for the sake of brevity, in the following explanation and discussion of the results we concentrate only on the effects of stretch of the muscles.

Figure 2. Typical records of the changes in fura-2 fluorescence ratio and force produced by shortening a rat trabecula by 10% for 15 min.

A, chart records of 340 nm/380 nm fluorescence ratio and force, with a representation of the length change from the initial length (L₀). A shutter in the excitation light pathway was opened only for discrete 48 s recording periods (labelled 1–5) in order to avoid photobleaching of fura-2. The shutter was closed also during the adjustment of muscle length using a micromanipulator. Traces were scanned digitally from the original records (filtered at 15 Hz). Note the slow changes in twitch force after the changes in muscle length. B, mean records (from 16 twitches) of fluorescence ratio and force measured during periods 1–5 in A. Unfiltered records. C, overlaid traces of the fluorescence ratio and force averaged during periods 3 (○) and 4 (•) to illustrate the rapid effects of the length increase. Resting forces have been subtracted from these traces. D, similar overlaid traces averaged during periods 4 (•) and 5 (⋄) to illustrate the delayed effects of the length increase. 24 °C, 1 mM external Ca²⁺, 0.33 Hz stimulation rate.

From Fig. 2 it is evident that the time courses of the twitch and the Ca²⁺ transient were also influenced by length. These changes are quantified in Fig. 4, which shows time-to-peak and the times for 50% relaxation (RT₅₀) and 90% relaxation (RT₉₀) of force and fluorescence. The changes in twitch time course followed closely the changes in twitch amplitude, in that muscle re-lengthening from L₉₀ to L₀ produced a rapid prolongation of the twitch time to peak, RT₅₀ and RT₉₀, followed by a further increase over 15 min (Fig. 4A, C and E). The Ca²⁺ transient, however, showed more complex alterations. Upon muscle re-lengthening, the Ca²⁺ transient time to peak decreased (the opposite of that for force), but stayed constant thereafter (Fig. 4B). At L₉₀ the decay of the Ca²⁺ transient was approximately exponential, but after stretching the muscle to L₀ there was a faster initial decay of the Ca²⁺ transient, followed by a slower second phase, which appeared as small ‘bump’ of Ca²⁺ superimposed on an exponential decay. This caused a crossover in the overlaid traces (Fig. 2C), as observed previously (Backx & ter Keurs, 1993). The bump of [Ca²⁺]_i was unlikely to be a movement artifact, since it occurred in opposite directions in the individual 340 and 380 nm signals (not shown), suggesting that it reflected a true change in [Ca²⁺]_i. These changes in the Ca²⁺ transient lead to a tendency for its RT₅₀ to fall, and for RT₉₀ to increase, immediately after muscle lengthening (Fig. 4D and F). However, the time course of the Ca²⁺ transient was not changed subsequently during the 15 min of the slow force responses.

One of the major aims of this study was to determine whether there were any changes in diastolic [Ca²⁺]_i that could, by altering SR Ca²⁺ loading, account for the observed changes in systolic [Ca²⁺]_i during the slow force responses. As Figs 2D and 3B show, we found no significant delayed increase in diastolic [Ca²⁺]_i after stretching these muscles. In fact, the slow increase of systolic [Ca²⁺]_i after muscle re-stretch was associated with a small fall of the diastolic ratio (from 0.481 ± 0.018 to 0.476 ± 0.018; P < 0.05, n = 15), although there was no significant slow change after muscle shortening. Thus the systolic changes were not associated with any similar changes in diastolic [Ca²⁺]_i. There were, however, changes in diastolic [Ca²⁺]_i immediately after the length changes, e.g. diastolic [Ca²⁺]_i fell significantly upon re-lengthening of the muscle (Fig. 3B).

Force and [Ca²⁺]_i in muscles with inactive SR

Although diastolic [Ca²⁺]_i did not appear to rise during the slow increase of force, it was possible that the Ca²⁺-sequestering activity of the SR could have masked any tendency for diastolic [Ca²⁺]_i to increase. For example, if there had been an extra influx of Ca²⁺ at longer lengths, this Ca²⁺ could have been immediately taken up by the SR and observed as an increase in systolic, but not diastolic, [Ca²⁺]_i. To investigate this possibility, we repeated the length changes after the muscles had been exposed to the SR inhibitors ryanodine and CPA for more than 1 h. This also allowed us to determine the contribution of the SR to the rapid and the slow force responses. As reported previously with single cells (Chiesi, Wrzosek & Grueninger, 1994; Lewartowski, Rózycka & Janiak, 1994), we found that adding CPA (30 μM) to trabeculae treated with ryanodine alone (1 μM) substantially increased the size of, and prolonged, the force and Ca²⁺ transients (results not shown); this indicates that in the presence of ryanodine alone the SR can take up Ca²⁺, and can trap some of the Ca²⁺ entering the cell via the sarcolemma (Chiesi et al. 1994; Lewartowski et al. 1994). We therefore used the combination of ryanodine (1 μM) and CPA (30 μM) to inhibit the SR more completely. Figure 5 shows results from a typical muscle and Fig. 6 gives the mean results. Both the twitch and the Ca²⁺ transients were reduced in magnitude and prolonged by SR inhibition (compare Figs 2 and 5). Nevertheless, the slow changes in force were still evident (Figs 5 and 6), and in fact were of the same magnitude as with a functional SR. For example, after muscle re-lengthening to L₀ the slow force increase was 38.9 ± 8.5% before SR inhibition and 31.5 ± 8.9% during SR inhibition (P > 0.05, paired t test, n = 6). The rapid changes in force were also unaffected by SR inhibition: muscle re-lengthening increased twitch force immediately by 149.2 ± 30.2%, compared with 144.1 ± 13.1% in muscles with functional SR (P > 0.05, n = 6). This suggests that the SR does not play a significant part in the rapid or the slow force responses to a length change.

Figure 5. Typical records of the changes in fura-2 fluorescence ratio and force produced by shortening a trabecula for 15 min in the presence of SR inhibitors.

A, chart records of 340 nm/380 nm fluorescence ratio and force, with a representation of the length change. The record was taken > 1 h after the addition of ryanodine (1 μM) and cyclopiazonic acid (30 μM). Other details as in Fig. 2. Note that force is lower than in Fig. 2, but the slow changes of force after a change of muscle length are still present. B, mean records (from 16 twitches) of fluorescence ratio and force measured during periods 1–5 in A.

Figure 6. The effects of the length changes on twitch force and on the diastolic and systolic fluorescence ratios in the presence of SR inhibitors.

A, twitch (active) force during the different periods (1–5) of Fig. 5A. B, systolic fluorescence ratio (○) and diastolic ratio (▵) during the same periods. Note the break in the ordinate. Muscles were incubated in the presence of ryanodine (1 μM) and cyclopiazonic acid (30 μM) for at least 1 h before measurements were taken. The abscissae indicate muscle length and the time in minutes at which data recording was started after the length change. L₉₀ represents a length of 90%L₀. Points show means ±s.e.m., n = 6. Where no error bars are shown s.e.m. is less than the size of the symbol. Symbols next to the data points indicate results of paired t tests between each value and the corresponding value in the preceding period. *P < 0.05; †P < 0.01; ‡P < 0.001.

With the SR inhibited, there were slow changes in systolic [Ca²⁺]_i after muscle shortening or re-lengthening, but there were still no time-dependent changes in the diastolic [Ca²⁺]_i (Fig. 6B). Thus there was no evidence that the SR activity had been masking any true change in diastolic [Ca²⁺]_i. There were, however, changes in diastolic [Ca²⁺]_i immediately after the length changes (Fig. 6B), as was found with the SR functional (Fig. 3B), e.g. diastolic [Ca²⁺]_i fell upon muscle re-lengthening.

Inhibition of the SR did, however, produce some differences compared with the results obtained with the SR functional. The first was that the slow responses were prolonged when the SR was inhibited: the half-time for recovery of force after muscle re-lengthening was 3.7 ± 0.2 min (n = 6) in the control muscles and 6.0 ± 0.8 min during SR inhibition. Second, we saw changes in the time course of the twitch similar to those in Fig. 2, but the time course of the Ca²⁺ transient, which exhibited a simple exponential decay with no bump, was not affected immediately after the length change (illustrated in Fig. 5; mean data not shown). As when the SR was functional, there were no time-dependent changes of Ca²⁺ transient time course.

Unstimulated muscles

Evidence has previously been presented (Nichols, 1985; Allen et al. 1988) that the mechanism responsible for the slow force response occurs during the diastolic period. Accordingly, in some experiments we rested the muscles to try to maximize any changes of [Ca²⁺]_i occurring during diastole (Fig. 7). The resting [Ca²⁺]_i in the unstimulated muscles was lower than during stimulation, whether the SR was functional (compare Figs 3B and 7A) or not (compare Figs 6B and 7B). This shows that our inability to detect a change in diastolic [Ca²⁺]_i in the stimulated muscles had not been due to the minimum fluorescence ratio being reached. In the resting muscles there was no significant rapid change in diastolic [Ca²⁺]_i. More importantly, once again we could detect no slow changes in diastolic [Ca²⁺]_i, whether SR inhibitors were absent (Fig. 7A) or present (Fig. 7B). Thus under no type of experimental condition did we detect any slow changes in diastolic [Ca²⁺]_i that could have lead indirectly to the observed slow changes in systolic [Ca²⁺]_i.

Variation of [Ca²⁺]_i during changes of [Ca²⁺]_o

To check whether our experimental procedure was sensitive enough to record any changes in diastolic [Ca²⁺]_i that were responsible for the observed slow changes in systolic [Ca²⁺]_i, we varied external [Ca²⁺] ([Ca²⁺]_o), which has been reported to alter systolic force and [Ca²⁺]_i by effects on diastolic [Ca²⁺]_i (Frampton, Orchard & Boyett, 1991). When [Ca²⁺]_o was decreased from 1 to 0.5 mM, which reduced twitch force to 42.0 ± 9.7% of control (n = 4), there was a small but non-significant decrease of the diastolic fluorescence ratio in stimulated muscles (to 98.2 ± 2.1% of control, n = 4) and a significant decrease (to 94.7 ± 1.9%, n = 3; P < 0.05) in resting muscles. During SR inhibition by ryanodine and CPA, these changes were greater, with the diastolic fluorescence ratio being reduced to 83.8 ± 4.7% (n = 4; P < 0.05) and resting ratio to 80.1 ± 4.2% (n = 4; P < 0.05). Thus changing [Ca²⁺]_o did have an effect on force that was associated with measurable changes in diastolic [Ca²⁺]_i. This suggests that, had changes in diastolic [Ca²⁺]_i been responsible for the slow force changes, we should have been able to detect them. Furthermore, the finding that [Ca²⁺]_o had a greater effect on [Ca²⁺]_i once the SR had been inhibited indicates that in the absence of the SR inhibitors the SR is likely to have been buffering diastolic [Ca²⁺]_i in the length-change experiments.

Length dependence of the myofibrillar force-[Ca²⁺]_i relationship

In Fig. 3 the large increase in twitch force (58 ± 7.5%, n = 15) during the slow response to muscle stretch was associated with only a small rise (10.9 ± 2.5%) in the magnitude of the Ca²⁺ transient. One question, therefore, is whether this change in the Ca²⁺ transient is large enough to account for the potentiation of force, or whether there is in addition an increased sensitivity of the myofibrils to Ca²⁺. We employed two approaches to answer this. In the first (Fig. 8), we compared the changes in force and [Ca²⁺]_i during the slow responses with those changes caused by alteration of stimulation frequency, when force is thought to be altered by variations in [Ca²⁺]_i alone (e.g. Frampton et al. 1991). Note that in our rat trabeculae the force-frequency relationship is positive. The relationship between peak force and peak [Ca²⁺]_i in the twitch was the same during the slow force responses as that produced by a change in frequency (Fig. 8). This suggests that the slow force responses may be explained entirely by the changes in systolic [Ca²⁺]_i.

Figure 8. Changes in force and systolic fura-2 fluorescence in a typical trabecula in response to alterations of stimulation frequency or muscle length.

The relationship between peak force and peak fluorescence ratio during the twitches was determined under two conditions: (i) the trabecula was held at constant length (L₀) and was stimulated at various frequencies from 0.2 to 2.5 Hz (▪); (ii) the muscle was stimulated at 1 Hz throughout and muscle length was reduced by 10% for 15 min (□). The numbers correspond to the periods shown in Fig. 1A. The slow force increase after re-lengthening of the muscle, i.e. from periods 4 to 5, exhibits the same force-[Ca²⁺]_i relationship as obtained with the changes of frequency. Similar results were seen in three other muscles.

Our second approach was to measure the myofibrillar force-[Ca²⁺]_i relationship directly. Backx, Gao, Azan-Backx & Marban (1995) showed that, if twitch relaxation has been slowed by inhibition of the SR, then [Ca²⁺]_i and force come into equilibrium during relaxation, and a plot of force against fluorescence ratio during relaxation reflects the steady-state force-[Ca²⁺]_i relationship of the myofibrils in the muscle. This type of plot (‘phase-plane plot’) is shown in Fig. 9 for twitches from a typical muscle before and after it was re-lengthened to L₀. The left-hand side of the anticlockwise loops represents the lower portion of the sigmoidal myofibrillar force-[Ca²⁺]_i relationship. After the muscle was stretched, the curve shifted to the left, corresponding to an increase in myofibrillar responsiveness to Ca²⁺. Over the next 15 min both systolic [Ca²⁺]_i and force increased, but the steady-state force-[Ca²⁺]_i relationship of the myofibrils was unaffected. A decrease in length produced similar effects, but in the opposite direction (not shown). Thus all the slow force changes were due to alterations in the magnitude of the Ca²⁺ transient, since there was no evidence for a contribution from a change in myofibrillar responsiveness to Ca²⁺.

Figure 9. Representative plots of force against fluorescence ratio measured throughout the twitches at different lengths.

The muscles was treated with ryanodine (1 μM) and cyclopiazonic acid (30 μM). The changes in force and fluorescence ratio throughout the twitch are plotted for twitches: in the steady state at L₉₀ (○), in the first minute after muscle re-lengthening to L₀ (•), and after 15 min at L₀ (⋄). Arrows indicate the direction of the loops. Note that the relaxation (left-hand) part of these phase-plane plots is shifted leftwards by the increase in muscle length, suggesting an increase of myofibrillar Ca²⁺ sensitivity, but then follows a constant trajectory over the time course of the slow rise of twitch force. Similar results were seen in three other muscles.

The apparent leftward shift in Fig. 9 immediately after stretch of the muscle could have been due to an increase of either the Ca²⁺ sensitivity or the maximum force production of the myofibrils. To assess the relative contribution of these two factors, we determined the entire force-[Ca²⁺]_i relationship of the myofibrils, using tetani to achieve maximum activation of the myofibrils (Backx et al. 1995). Figure 10A and B shows fluorescence and force, respectively, during tetani elicited in a muscle at L₀ and L₉₀. During tetanization, force reached a stable value in 1–2 s, whereas [Ca²⁺]_i continued to rise until the stimulation was terminated after 3 s; this indicates that full Ca²⁺ activation of the myofilaments was achieved during the tetanus. In spite of this, tetanic force was decreased by more than 25% by the 10% shortening of the muscle. The force-[Ca²⁺] relationship of the myofibrils (Fig. 10C) was obtained from the slow fall of [Ca²⁺]_i and force after the end of the train of stimuli. The curves were well fitted by the Hill equation (see legend of Fig. 10) and were found to be steep, with an n_H value of 7.57 at L₀. (Note that this n_H parameter is different from that normally used to describe force-[Ca²⁺] relationships, because here we used fluorescence ratio rather than [Ca²⁺].) In addition to the decrease of maximum force in the shortened muscle, there was a decrease of Ca²⁺ sensitivity, shown by the increase in K_½ from 2.03 to 2.59, and a slight decrease in n_H. Similar results were seen in two other muscles. Figure 10D presents results from three experiments where various length changes were imposed on the muscles. The maximum Ca²⁺-activated force of the myofibrils fell substantially at shorter lengths, with the standard 10% shortening of the muscle decreasing tetanic force to 63.1 ± 3.2% of that at L₀. Thus muscle length influenced both the maximum force production and Ca²⁺ sensitivity of the myofibrils in situ in the trabeculae, and both factors were likely to have contributed to the rapid changes in twitch force when length was altered.

Figure 10. Force and fluorescence ratio during tetanic contractions at various muscle lengths.

A and B, fluorescence ratio and force, respectively, during tetanic contractions at L₀ (•) and at L₉₀ (○) in a typical muscle. Tetani were produced by 10 Hz stimulation (in 8 mM Ca²⁺, 1 μM ryanodine and 30 μM cyclopiazonic acid). C, force versus fluorescence ratio during the relaxation phase of the tetani at L₀ (•) and L₉₀ (○). Sigmoid curves are fits to the Hill equation: force = maximum force × ratio^n_H/(K_½ⁿ_H+ ratio^n_H), with n_H= 8.05 and K_½= 2.03 at L₀ and with n_H= 7.57 and K_½= 2.59 at L₉₀. D, tetanic force at different muscle lengths. Forces are expressed relative to that at L₀. Means ±s.e.m. of three experiments. Where no error bars are shown, s.e.m. is less than the size of the symbol.

One other interesting observation was that the rise of [Ca²⁺]_i during the tetanus was also length dependent (Fig. 10A), with the [Ca²⁺]_i measured at 2–3 s being elevated at longer length. In three muscles the fluorescence ratio 3 s after the start of tetanic stimulation was 19.0 ± 2.9% higher at L₀ than at L₉₀ (P < 0.05), even though a difference was not apparent with the first stimulus of the train. This suggests that, during frequent action potentials, the loading of the cell with Ca²⁺ is length dependent (see Discussion).

DISCUSSION

Rapid changes of force and [Ca²⁺]_i after a length change

Changes of muscle length produced rapid, large changes in the magnitude of the twitch, but no apparent change in the magnitude of the Ca²⁺ transient, whether the SR was functional (Figs 2 and 3) or not (Figs 5 and 6); this suggested that the changes in force resulted predominantly from changes in the myofibrillar force-[Ca²⁺]_i relationship. An increase in myofibrillar Ca²⁺ sensitivity at longer length has previously been inferred from indirect evidence in intact preparations (e.g. Allen & Kurihara, 1982), and has been shown directly in skinned trabeculae (Hibberd & Jewell, 1982; Kentish, ter Keurs, Ricciardi, Bucx & Noble, 1986). Here we show, for the first time, that the steady-state force-[Ca²⁺]_i relationship of the myofibrils in situ in the muscles is indeed changed; this was established by plotting the force-fluorescence loops from twitches (Fig. 9) or tetani (Fig. 10) in muscles in which the SR was inhibited (in order to allow force and [Ca²⁺]_i to come into equilibrium during relaxation: Backx et al. 1995). After muscle stretch there was an immediate increase in the myofibrillar responsiveness to Ca²⁺ (Fig. 9), which was found to be due to three factors (Fig. 10C): an increase of myofibrillar Ca²⁺ sensitivity; an increase in the maximum force production (at saturating [Ca²⁺]_i); and a small increase in the slope of the force-[Ca²⁺]_i relationship. These changes were similar to those in skinned muscle (Kentish et al. 1986). The increase of Ca²⁺ sensitivity at longer lengths is thought to be due to a rise in the affinity of troponin C (TnC) for Ca²⁺ as a consequence of an increase in the number of force-generating cross-bridges (Bremel & Weber, 1972; Allen & Kentish, 1985, 1988; Hofmann & Fuchs, 1988; Kurihara & Komukai, 1995); this may result from a decrease in myofilament lattice spacing (McDonald & Moss, 1995). The potentiation of maximum force is probably the result of less double overlap of myofilaments and a smaller restoring force arising from the extracellular tissue matrix (Kentish & Stienen, 1994). The greater slope may be a consequence of greater co-operativity between cross-bridges at longer sarcomere lengths (Kentish et al. 1986). Of these three factors, the change in Ca²⁺ sensitivity accounts for most of the rapid change in twitch force as length is increased. From data such as those in Fig. 10 we conclude that, at the degree of Ca²⁺ activation of the myofibrils during the twitch at L_o (∼30% of maximum activation from a comparison between Figs 3 and 10C), about 60% of the threefold increase in twitch force after muscle stretch (Fig. 3A) is due to the increase in Ca²⁺ sensitivity, while the remaining 40% is due chiefly to the increase in maximum force production, with the change in slope making only a minor contribution.

An alteration of myofibrillar Ca²⁺ sensitivity may also explain the length-dependent changes in the time course of the Ca²⁺ transient. As found previously using aequorin (Allen & Kurihara, 1982), at longer lengths the early decay of the Ca²⁺ transient is accelerated (Figs 2C and 4D), which can be explained by a potentiated binding of Ca²⁺ to TnC, causing the cytosolic [Ca²⁺]_i to fall faster because the SR and other Ca²⁺ removal processes can lower it more easily. On the other hand, the second half of the Ca²⁺ transient decay was prolonged at longer length, as a result of the appearance of the ‘bump’ of [Ca²⁺]_i (Figs 2C and 4F). The bump of [Ca²⁺]_i could be due to slower dissociation of Ca²⁺ from TnC, or to SR Ca²⁺ release at longer lengths (Backx & ter Keurs, 1993). It could also be due to accelerated Ca²⁺ release from TnC, prompted by the co-ordinated detachment of myosin cross-bridges from actin during muscle relaxation, which would reduce the Ca²⁺ affinity of TnC (see above). At longer lengths, relaxation occurs later (Fig. 2C), so Ca²⁺ release from TnC would be delayed.

Our finding that the rapid change in twitch force after the length change was not significantly affected by the presence of SR inhibitors ryanodine and CPA confirms that the SR contributes little, if anything, to the rapid effects of a length change. Although we cannot be sure that SR activity was abolished completely under these conditions, it is likely that SR activity was reduced substantially (discussed below), so if the SR had been contributing significantly to the rapid force change we should have seen a depression of its magnitude.

The only rapid effects of length on [Ca²⁺]_i were the apparent increases and decreases of diastolic [Ca²⁺]_i when muscle length was reduced and increased, respectively (Figs 3 and 6). These changes (which may if anything have been slightly underestimated by our calculation of the 340 nm/380 nm ratio using the autofluorescences measured at L₀; see Methods) are in the opposite direction to explain any slow changes in systolic [Ca²⁺]_i. They may arise from the reduced Ca²⁺ binding to TnC at short muscle lengths, since there is some binding of Ca²⁺ to TnC even at Ca²⁺concentrations where no force is developed (e.g. Hofmann & Fuchs, 1988).

Slow changes of force and [Ca²⁺]_i after a length change

It is known that the slow force responses are associated in time with changes in the magnitude of the Ca²⁺ transient, measured in preparations injected with aequorin (e.g. Allen & Kurihara, 1982) or loaded with fura-2 AM (Steele & Smith, 1993; Hongo et al. 1996). However, it has not previously been established whether the changes of the Ca²⁺ transient are great enough to account quantitatively for the changes in force. Indeed, the slow changes of the Ca²⁺ transient measured with fura-2 (Hongo et al. 1996; present results), or of SR Ca²⁺ content measured with rapid cooling contractions (Bluhm & Lew, 1995), are generally only 20–30% of the slow changes in twitch force (e.g. compare Fig. 3A and B). Thus, it seemed possible that some of the slow force response could have been due to an increase in myofibrillar responsiveness to [Ca²⁺]. However, the concurrence of the relationships between peak force and peak [Ca²⁺]_i during the slow force responses and after changes in frequency (when force is probably altered only by changes in the Ca²⁺ transient: Frampton et al. 1991), suggested that the slow force responses were due only to the changes in [Ca²⁺]_i (Fig. 8). This was confirmed by the demonstration that the steady-state force-[Ca²⁺]_i relationship of the myofibrils was the same at the beginning and at the end of the slow force responses (Fig. 9). Thus all the slow force response was indeed due to the change in Ca²⁺ transient. The small increments in systolic [Ca²⁺]_i produced large slow force responses because the force-[Ca²⁺] relationship of the myofibrils in situ was steep (Figs 9 and 10).

We demonstrate here that the slow responses in force and [Ca²⁺]_i have an undiminished magnitude under conditions designed to inhibit the SR. A lack of effect of ryanodine on the slow force responses has been reported previously (Kentish, Davey & Largen, 1992; Bluhm & Lew, 1995; Hongo et al. 1995). However, some SR Ca²⁺ uptake can still occur in the presence of ryanodine (Lewartowski et al. 1994; Chiesi et al. 1994; J. C. Kentish & A. Wrzosek, unpublished observations), so we added CPA to reduce Ca²⁺ uptake further. It is likely that the SR was inhibited substantially, if not completely, under these conditions, because (i) twitch force was reduced by ∼70% (compare Figs 3A and 6A), (ii) CPA had an inhibitory effect (increase and prolongation of Ca²⁺ transient) additional to that of ryanodine, and (iii) Baudet, Shaoulian & Bers (1993) have reported that ryanodine (10 μM) plus CPA (100 μM) abolishes SR Ca²⁺ loading in rabbit muscles. Here we establish (Figs 5 and 6) that the presence of SR inhibitors did not diminish the slow changes of force or of the Ca²⁺ transient. These results indicate that the fundamental process responsible for the slow responses does not depend upon the SR. Interestingly, although SR inhibition did not affect the magnitude of the slow force responses, it did slow their rate of development, suggesting that the SR may in some way accelerate the process underlying the slow force response.

What mechanism is then responsible for the slow changes in twitch force and in systolic [Ca²⁺]_i? Long-term changes in contractility of cardiac cells are likely to be due to changes in Ca²⁺ fluxes across the cell membrane. Work by Nichols (1985) and Allen et al. (1988) had indicated that it was diastolic, rather than systolic, length that seemed to be the important variable, from which it seemed likely that a slow change of diastolic [Ca²⁺]_i was responsible for the slow changes of systolic [Ca²⁺]_i. Indeed, Steele & Smith (1993) did find appropriate changes in diastolic [Ca²⁺]_i in isolated guinea-pig trabeculae loaded with fura-2 AM. However, we detected no time-dependent changes of diastolic [Ca²⁺]_i after muscle lengthening (cf. Hongo et al. 1996), or after muscle shortening (Figs 2 and 3). Furthermore, we found that slow changes of diastolic [Ca²⁺]_i were not seen in unstimulated muscles (Fig. 7), when the diastolic period is in effect maximal. However, it was possible that the highly active SR in rat myocardium had obscured any true changes of diastolic [Ca²⁺]_i, since the effects on [Ca²⁺]_i of reducing [Ca²⁺]_o from 1.0 to 0.5 mM were much larger after SR inhibition. However, even after the SR had been inhibited we could detect no slow changes of diastolic [Ca²⁺]_i in stimulated or unstimulated muscles (Figs 5–7). Our data, therefore, indicate that in rat myocardial tissue there is no apparent change in diastolic [Ca²⁺]_i that can account for the slow changes in systolic [Ca²⁺]_i and force.

Our conclusion that diastolic [Ca²⁺]_i does not underlie the slow force responses differs from that of Nichols (1985) and Allen et al. (1988), and is not consistent with the suggestion (Allen et al. 1988) that the slow responses are due to activation of stretch-activated channels, leading to an increased diastolic [Ca²⁺]_i. There could, however, be diastolic changes in other cellular factors that alter the systolic Ca²⁺ transient, such as cAMP. It was suggested that cellular levels of cAMP may be length dependent, since a maintained elevation of cAMP levels produced by isoprenaline blocks the slow force responses (Kentish et al. 1992). In addition, Singh (1982) reported a stretch-induced increase in [cAMP] in frog ventricular muscles. On the other hand, a rise in [cAMP] after muscle re-lengthening should promote SR Ca²⁺ uptake (via protein kinase A-mediated phosphorylation of phospholamban), resulting in a pronounced shortening of the twitch. In fact, the twitch was significantly prolonged during the slow force increase (Fig. 4). Thus an essential role for a rise of [cAMP] in the slow responses seems unlikely, unless this cAMP is localized to a sub-sarcolemmal space. Another possible mediator of changes in the Ca²⁺ transient is inositol trisphosphate, which increases in concentration after myocardial stretch (Dassouli, Sulpice, Roux & Crozatier, 1993).

The simplest explanation for our results is that the slow changes in systolic [Ca²⁺]_i are due to a change in trans-sarcolemmal Ca²⁺ movements during systole, such that cellular Ca²⁺ loading is enhanced at longer muscle lengths. A greater Ca²⁺ influx could result from an increase in I_Ca, from a reduced Ca²⁺ extrusion during systole, or indirectly from prolongation of the action potential. In this regard, Allen (1977) reported that the action potential in cat papillary muscles was prolonged by muscle stretch. In each case, a length dependence could be due to a direct influence of cell length (membrane stretch) on the channel, or to a localized increase of cAMP, etc., leading to increased channel flux via phosphorylation of the channel. These possibilities remain to be investigated. Although a change in I_Ca seems the simplest possibility, Hongo et al. (1996) found no slow change in I_Ca. Unfortunately their results were not conclusive because in the cells where I_Ca was measured there were also no slow force responses. One pertinent result from our experiments is that during the tetani there was no apparent difference in [Ca²⁺]_i between long and short lengths with the first Ca²⁺ transient, but the subsequent rise of [Ca²⁺]_i during the tetanus was greater at the longer length (Fig. 10). This suggests that an increase of length does indeed increase the cellular loading of [Ca²⁺]_i in a time-dependent fashion. The rapid development of the length-dependent divergence in [Ca²⁺]_i during tetani, when action potentials are frequent, is consistent with the idea that the alteration of Ca²⁺ loading occurs during the action potential itself. However, the precise basis of this length and time dependence of [Ca²⁺]_i remains to be elucidated.

In conclusion, we find that the rapid alteration of twitch force after a length change in rat trabeculae is due chiefly to the alteration of myofibrillar Ca²⁺ sensitivity and, to a lesser extent, to a change in the maximum force production of the myofibrils. In contrast, the delayed change in twitch force is not due to a change in myofibrillar responsiveness to Ca²⁺, but is due entirely to changes in the magnitude of the Ca²⁺ transient. Whether the SR is active or not, these slow changes of the Ca²⁺ transient occur without any corresponding change in the diastolic [Ca²⁺]_i. Our evidence suggests there is a length- and time-dependent change in the Ca²⁺ loading of the cells during the systolic period, probably during the action potential itself.